Topical Dutasteride: The 2025 Trial's Numbers Don't Add Up
Published by AmpleLab Research
In 2025, a randomised, double-blind Phase II trial reported that a 0.05% topical dutasteride solution significantly outperformed oral finasteride at growing hair, with minimal change to serum DHT. To our knowledge, it's the first daily, standalone topical dutasteride trial ever run without microneedling, and the headline has since circulated widely as proof that topical dutasteride has arrived.
A close read of the trial's own published numbers tells a messier story. Three separate, independently checkable details in this paper don't line up with what you'd expect from a clean result: the starting hair counts, the dose-response pattern, and the comparator arm's own hormone readings. None of them individually proves the headline wrong. Together, they're enough that nobody should be citing this trial as settled evidence yet.
To be clear about scope: this isn't a case against topical dutasteride as an approach. Earlier, smaller studies pairing it with microneedling have reported modest, believable results consistent with how the drug is understood to work, covered in our topical vs oral dutasteride article. The concern here is specific to this one trial and its claim to outperform oral finasteride outright, which is a different and much larger claim than the evidence base for topical dutasteride generally supports.
The Trial Behind the Claim
A multicentre, randomised, double-blind, double-dummy, placebo- and active-controlled Phase II trial across six sites in India (CTRI/2021/03/032173) enrolled 135 men aged 20-60 with Norwood III-vertex, IV, or V androgenetic alopecia. Participants were split 2:2:2:2:1 into topical dutasteride 0.01%, 0.02%, or 0.05%, oral finasteride 1mg, or placebo, for 24 weeks. Every participant received both a topical and an oral product (active or matched placebo in each) to preserve blinding across both routes. The primary endpoint was target area hair count (TAHC) at week 24, measured in a defined area at the vertex using Dino-Lite digital microscopy after clipping hairs to 0.5-1mm and marking the centre of the measurement zone.
The trial was funded by Shilpa Medicare Limited, the company that developed the formulation being tested. All seven listed authors are affiliated with Shilpa Medicare Limited; there is no independent academic co-author on the paper, and no formal sample size calculation was performed. None of that means the results are fabricated. It does mean this is industry-funded, industry-authored data from a single trial without an independent academic co-author, which is exactly the kind of result that warrants a closer look before being treated as settled evidence.
What the Trial Reported
At week 24, mean change in non-vellus hair count came out at +32.32 hairs/cm² for 0.01% dutasteride, +27.48 for 0.02%, +75.52 for 0.05%, +41.21 for oral finasteride, and +0.07 for placebo. The 0.05% dutasteride group's improvement over finasteride reached statistical significance (p=0.0083), the specific comparison behind the "beats finasteride" headline. Hair width improved across all active arms (0.01%: +6.68μm, 0.02%: +9.15μm, 0.05%: +11.59μm, finasteride: +10.68μm, placebo: +4.00μm), though only 0.05% and finasteride beat placebo significantly here, and 0.05% did not significantly outperform finasteride on width specifically.
Investigator-rated global photographic assessment (GPA) scores of "moderate improvement or better" came in at 68.97% for 0.05% versus 21.43% for finasteride and 13.33% for placebo. Patient-reported satisfaction followed the same pattern: 96.55% for 0.05% versus 71.43% for finasteride and 33.33% for placebo. No serious adverse events were reported in any arm, and skin irritation occurred at similar low rates across active groups. Taken at face value, this is an unusually strong result for a topical product going up against an established oral drug. The three problems below are why it shouldn't be taken at face value yet.
Problem One: A Baseline That Doesn't Match Comparable Trials
The trial reported baseline non-vellus hair density of roughly 305 to 330 hairs/cm² across its treatment groups, before any treatment had been applied. Published phototrichogram studies using comparable scalp-counting methods generally report lower densities. A frequently cited reference study of normal adult scalp found a mean of 223 hairs/cm² (range 175-300)¹, and a more recent regional study of healthy adults found a mean of roughly 147 hairs/cm² at the vertex². In men specifically diagnosed with androgenetic alopecia, a 2024 cross-sectional study reported mean vertex density falling from around 149 hairs/cm² in unaffected controls to around 122 hairs/cm² at the most advanced stage measured, with individual values as low as 46³. None of these studies reports anything close to this trial's 305-330 hairs/cm² baseline, in either balding or non-balding scalp.
This isn't a comparison against a single outside opinion. A separate, unrelated Phase 2 trial of a different compound enrolled a similarly Norwood-staged population and reported baseline non-vellus hair counts roughly a third of what this trial reported, for men with comparable hair loss severity:
A multicentre, randomised, double-blind, placebo-controlled Phase II trial of GT20029 screened 225 men and enrolled 180 with Hamilton-Norwood grade IIIv to V androgenetic alopecia across 12 sites in China, all of Han Chinese ethnicity. At baseline, the average non-vellus hair count across groups sat between roughly 105 and 120 hairs/cm², before any treatment was applied.
Two independently run trials landing nearly threefold apart on baseline density, before either treatment did anything, raises a real question about a measurement or definitional difference between the two. Worth noting for fairness: the GT20029 trial enrolled a Han Chinese population in China, while this trial enrolled participants in India, and population-level differences in baseline hair density across ethnic groups are documented in the literature. That alone doesn't close a threefold gap, but it means the discrepancy isn't necessarily down to methodology or data quality; it could partly reflect who was being measured, not just how.
The published paper doesn't specify a hair-diameter cutoff for distinguishing non-vellus (thicker, pigmented) hairs from vellus (fine, miniaturised) ones, a detail other trials in this space routinely specify. For comparison, a well-known oral dutasteride Phase 3 trial (NCT01231607) explicitly defined non-vellus hair as ≥30μm in diameter and used a cosmetic ink tattoo dot to mark the exact centre of its measurement circle, redone if it faded, specifically so the same patch of scalp could be relocated at every visit. This trial describes neither a diameter cutoff nor a tattoo or template method, only that the centre point was marked at baseline. Either an absent diameter threshold, letting thin vellus hairs count alongside terminal ones, or a measurement area that drifted between visits, could explain a baseline this far outside the expected range. The paper doesn't give readers enough detail to rule either out.
Problem Two: A Dose-Response That Skips a Step
The paper describes its hair count results as dose-dependent. The 0.05% arm supports that framing on its own, but the three doses together don't: the 0.02% arm (+27.48 hairs/cm²) performed worse than the 0.01% arm (+32.32 hairs/cm²), despite containing twice the concentration of active drug. A clean dose-response would generally be expected to trend upward as concentration increases, not dip in the middle before jumping sharply at the top dose.
Small trial arms (30 patients each) can produce noisy, non-monotonic results by chance alone, so this isn't disqualifying by itself. But it's a second independent sign, on top of the baseline figures, that this trial's numbers hold together less cleanly than the topline comparison against finasteride suggests.
Problem Three: The Comparator Drug Didn't Behave Like Itself
This is the one most worth sitting with, because it doesn't depend on any comparison outside the trial itself. Oral finasteride 1mg is one of the most thoroughly characterised drugs in this entire field, and its dose-response on serum DHT was established decades ago:
A placebo-controlled, dose-ranging study measured scalp and serum androgen levels in men with androgenetic alopecia treated with finasteride at 0.01mg, 0.05mg, 0.2mg, 1mg, or 5mg daily for 42 days. Serum DHT declined significantly at every active dose, by 49.5%, 68.6%, 71.4%, and 72.2% at the 0.05mg, 0.2mg, 1mg, and 5mg doses respectively, confirming that the standard 1mg dose reliably suppresses serum DHT by roughly 70%, essentially matching the 5mg dose.
That finding has been reproduced in numerous independent studies since, and the suppression holds steady with continued use rather than fading.
This trial's own finasteride arm didn't reproduce that. The paper's own text states: "for the oral finasteride group, DHT levels decreased by 27% at week 12 but showed an incremental increase to an 11% reduction by week 24." In other words, the standard dose of one of the best-characterised drugs in dermatology produced roughly a third of its expected effect, and that effect then weakened over the course of the trial rather than holding steady. The dose given was correct; 1mg is the standard prescribed dose. The measured result from it was not what thirty years of independent research says it should have been.
A trial whose own positive control doesn't behave the way that drug reliably behaves elsewhere raises a direct question about the trial's hormone measurements, sample handling, adherence, or other study-specific factors. It's an internal reason to be cautious about the other measurements used to support the headline claim.
Where This Leaves Topical Dutasteride
None of this means topical dutasteride doesn't work, and none of it means this specific trial's authors manipulated anything. What it means is that three independent, checkable details in one paper, an implausible baseline, a dose-response that doesn't trend cleanly, and a gold-standard comparator drug behaving nothing like it does everywhere else, all point the same direction: treat this trial's magnitude of effect as unconfirmed, not as the settled word on whether topical dutasteride beats oral finasteride.
Two earlier, smaller studies pairing low-dose topical dutasteride with microneedling (2018 and 2022) found modest, believable effects with minimal systemic hormonal impact, consistent with dutasteride's known pharmacology. That modest picture is currently a more defensible baseline for what topical dutasteride may achieve. This trial's claim to have beaten oral finasteride outright, at a standalone dose with no microneedling, is the extraordinary part, and extraordinary results from a single, company-funded, unreplicated trial with this many internal inconsistencies need independent confirmation before anyone treats them as fact.
Frequently Asked Questions
Does topical dutasteride outperform oral finasteride?
One 2025 trial reported that 0.05% topical dutasteride significantly beat oral finasteride on hair count. That same trial's baseline hair counts are far higher than comparable trials report, its dose-response pattern skips a step, and its finasteride arm showed roughly a third of finasteride's well-documented DHT-suppressing effect. Treat the headline as unconfirmed, not settled.
Why does it matter that the finasteride arm looked off?
Finasteride's DHT-suppressing effect (65-70%) is one of the most consistently replicated findings in this entire field. This trial's own finasteride arm showed only 11-27%, and that effect declined rather than held steady over 24 weeks. A positive control behaving that differently from its established pharmacology raises a direct question about the trial's own hormone measurements and other study-specific factors.
Does this mean topical dutasteride doesn't work?
No. Two smaller, earlier studies pairing low-dose topical dutasteride with microneedling found modest, believable effects with minimal systemic impact. The issue is specifically with this one trial's claim to beat oral finasteride outright, which rests on data containing several unexplained inconsistencies.
Who funded this trial?
Shilpa Medicare Limited, the company that developed the topical dutasteride formulation being tested. All seven authors are affiliated with Shilpa Medicare Limited, with no independent academic co-author on the paper.
Selected Research
This article is provided for educational purposes and does not constitute medical advice. Topical dutasteride is not an FDA or MHRA-approved formulation and is available only through pharmacy compounding against an individual prescription. Consult a qualified healthcare professional before use. AmpleLab products are cosmetic formulations and are not intended to diagnose, treat, cure, or prevent any condition.
AmpleLab.